This is the content of a talk I gave today, which I recorded on my phone and then used AI to transcribe. I’ve given it a very quick look over and tidy up, but errors of transcription may remain!
An overview of my arguments
I’m going to preface this by saying that I took the overnight bus from Austin to New Orleans, and I have the worst headache. So, the underlying argument I want to make is something like liberalism trades on the idea that individual humans have the capacity to bear knowledge. This relates to a different argument I made recently that liberalism trades on the idea that individual humans have the capacity for free agency.
These arguments relate, I think, because I find it hard to see how someone could do stuff like acting of their own accord if they didn’t have the capacity to do stuff like bearing knowledge. I’ll come on to some of that, but I’m going to try to make an easier argument today, which will bear on that bigger argument in some sense, which is that it’s costly to the coherence of liberalism when liberals cut away at the significance of the idea that individual humans have the capacity to bear knowledge.
I’m going to focus on two ways in which this is undercut. One, by cutting away at the significance of what individual humans are doing when they bear knowledge. And second, when people promote the idea that things aside from individual humans can bear knowledge — so, non-human things, even non-living things, non-individual things. Both of these two approaches, I think, cut away at the idea there’s something special, maybe something morally relevant, about individual humans being able to bear knowledge.
I’m going to focus on an example of each of these two ideas that I’m going to call ‘the surprising views’. This is because these are views that I sometimes hear liberals make, and it’s surprising to me when I hear liberals make them. The first is a view that focuses heavily on the idea of tacit knowledge. This is a view that at its extreme says that most, or even all, of the things that we individual humans know, we know in some kind of mystical, inarticulable way. And the second is a view that groups of things can bear knowledge, qua groups. I think perhaps the most surprising thing to me about these views being espoused by liberals is that they seem extremely costly to the individualism that is a part of liberalism.
Some context and caveats
The first bit of context I want to give is that I’m currently a bit obsessed by working out some kind of ontology of human capacities. So, going back to free agency, I believe that the capacity for free agency is contingent on the capacity to bear conscious mental states, which is contingent on the capacity to bear mental states. So, there’s some kind of relation between these kinds of capacities that I’m very interested in. I think about this in relation to AI at the moment quite a lot. And therefore, part of the value for me in this project about knowledge is working out where ‘bearing knowledge’ comes in, again, as a kind of matter of capacity. So, what kinds of things can ‘do knowing’? This hinges on the idea that knowledge is borne, that knowledge needs a bearer.
So, I’m sort of presupposing that. But I think we can get at this pretty ordinary idea by thinking, for instance, about the difference between knowledge and facts about the world. So, a book can contain facts about the world. We can argue about what that means, but I think you know what I’m getting at. This seems very different from saying that the book can bear knowledge. Bearing knowledge in the ordinary sense seems contingent on having mental states like thoughts, being the kind of thing that persists psychologically for long enough to count as knowing something, as opposed to maybe just, like, registering something or something like that. And these are properties we don’t usually attribute to books!
Second, I’m not really going to say much about what I take knowledge to be. This is obviously a very big philosophical problem — it pertains to the difference between knowledge and belief, for instance. But I am going to say that I’m referring to something stronger than just holding beliefs, or holding justified beliefs. In particular, I’m kind of talking about something that, like, tracks particular facts about the world in some sense. You can think of me as some kind of Nozickian-tracking-theory person, if you’re an epistemologist, and we can talk about that if you want. But, nonetheless, a lot of what I’m going to say also relates to holding beliefs as well as bearing knowledge. So, I think that’ll do.
Finally, I want to say that I could have picked various surprising views that liberals hold about knowledge to focus on, here, but I needed some examples that would enable me to talk about this risk of cutting away at the significance of the idea that individual humans bear knowledge to liberalism. And the two views that I’ve picked — this tacit knowledge view and the group knowledge view — I think serve me pretty well. What’s more, these two views are brought together in that I sometimes hear these views espoused by Hayekians, and this is useful for me. Partly because I enjoy thinking about Hayek, partly because Hayek was very interested in knowledge, and partly because Hayekians are often seen as particularly individualistic. So, this kind of adds to the surprisingness!
Some simple explanations in the context of Hayek
I’m going to begin by offering some simple explanations for these surprising views. And here I’m going to focus on the context that, at least in my experience, these views often arise within discussion of the knowledge argument, or discussions that are adjacent to this. ‘The knowledge argument’ is a name sometimes given to a theory Hayek advances in The Use of Knowledge in Society that broadly tells us something like: the price system arises from many individuals using the small bits of relevant on-the-spot knowledge that they each have to adjust their economic behaviour in response to other people adjusting their economic behavior, in such a way that enables them to trade goods with each other pretty efficiently.
So, this take of what Hayek is saying makes it into a descriptive theory. It’s Hayek saying something like, “Isn’t it amazing that this complex web of non-top-down directed interactions have these pretty great outcomes!?” Now, Hayek also says that a single central authority couldn’t get such great outcomes by, for instance, using clever statistical methods to work out prices — not only because the clever statistics couldn’t replicate all the little on-the-spot adjustments that each person makes relative to their own time and place, but also because the central authority couldn’t do these things for the people, or they wouldn’t be doing these things for themselves. But I don’t want to get into too many details about Hayek. I just want to emphasise that seeing the two surprising claims in this context can help us see that there might just be some pretty basic confusions going on.
Some simple explanations for the second surprising view
So first, when people who like the Hayekian knowledge argument say things like, “The group knows things that no member of the group knows,” or, “The market, as in the group of all the people, or even as in some non-living abstract system, knows all of these things,” they might simply be meaning something like the following. Nobody knows all the things, but when the individual knowers interact, they bring the relevant facts of the matter together. So, this could be analogous to someone saying, “You can find knowledge in books,” as a shorthand of saying, “There’s information in books that people can learn and come to know by reading them.”
Relatedly, it could simply be a case of people taking Hayek’s theory to tell us that because all of the bits of information are crucial, but these different bits are known by different people, then surely all of the bits can be added together to make ‘The Knowledge’. And then, because if you can’t have knowledge without a knowledge bearer, there must be some bearer of The Knowledge. And this bearer of the knowledge must be the group of the people, or the non-living system, or something else that is over and above the people who are involved, because the thing we do know is that it’s not any of them in particular who knows it all! Hayek is very clear about that. In other words, the knowledge is there, so someone or something must bear it. That could be an explanation.
Of course, alternatively, there is a familiar position in philosophy, which I’m sure you guys might go into, on which groups really can bear knowledge. But if the surprising view reflects that idea, then this seems totally against what Hayek himself is arguing in that paper, which is that we’re talking about a case in which there is no single knower. As he puts it, there is “no single mind”!
Now, at this point, you might just want to tell me that, look, the group is not a ‘single’, it’s a group. But here we get to the problem that the idea of ‘the group qua knowledge bearer’ is indeed the idea of the group as unified. This means that when you say the group knows, you could again just be referring to the individual knowers in the group each knowing bits. Like, when the group of us go to the beach, and John knows the first bit of the journey, and Oliver knows the second bit, and I know the third bit, none of us know all of it, but together we get to the beach. That doesn’t mean that the group of us knows some particular thing over and above each of the individual people knows! So, it could just be that which is happening.
But if what is meant when people say ‘the group knows’ is that the group bears knowledge above and beyond what the individual members of the group know — both in the sense that there is a group knower that is different from the individual knowers, and in the sense that the group knows stuff that none of the individuals know — then that is clearly at odds with the knowledge argument on my reading. And this is because, in the knowledge argument, Hayek is telling us explicitly there is no knower who knows all the things.
Some simple explanations for the first surprising view
Now, I’m going to turn to tacit knowledge. And again, I think there are some simple possible explanations for why it is that liberals will say sometimes that most or all the knowledge we hold is held tacitly. I experienced an example of this just the other day. A friend of mine, an Austrian economist, was talking about what he thinks knowledge to be. He made a nice distinction between information about stuff and knowing stuff. And then he basically admitted, no, most of the time, I think when we know stuff, we know it tacitly. This was very surprising to me.
Now, what might be going on in these instances — though I don’t think it is going on in the instance of my friend — is it might simply be that some people who like Hayek also like Polanyi, who is big on tacit knowledge, and they want to synthesise the key ideas of these people. So, all they’re doing is trying to bring Polanyi into Hayek. After all, people often have the desire to bring together the views of the people they admire. I think this would be a mistake, but it’d be understandable.
Second, Hayek does go off on a little rant in the paper about the wonders of human beings being able to do stuff without thinking about it, and of being able to use systems that we don’t understand. He says these things are very important for progress. The problem comes, however, if you think that this kind of mysteriousness applies to knowing those on-the-spot bits of knowledge that are crucial to the knowledge argument — the bits that each of us know relative to our time and place.
So, you could think here of the standard epistemological distinction between ‘knowing how’ and ‘knowing that’, on which I know that I took the bus from Austin to New Orleans last night, and I know how to ride a bike, but I can’t necessarily explain knowing how to ride a bike. Now, I think it’s crucial to Hayek’s argument that those little bits of knowledge that the different people hold in the knowledge argument is ‘knowledge that’. Each person knows things, and they act on things. Sure, this knowledge is limited, and sometimes it’s quite rough, in the sense of depending on some approximations. But it is good enough to contribute to the efficient system, because each person knows some relevant stuff well enough to reason on it and act on it. And this involves them doing informed things like bartering, spending extra, hiring the marginal worker.
So, this leads me to think, therefore, that the surprising commitment — over-commitment! — to tacit knowledge that I’ve sometimes come across may simply be a confusion between ‘limited knowledge’ in the sense of not being able to articulate the ‘knowing how’ of riding the bike, and ‘limited knowledge’ in the sense of only having small or rough bits of ‘knowing that’. I should, of course, note that ‘tacit knowledge’ isn’t always presented as being the same as ‘knowing how’. But nonetheless, it does always have this similar kind of hard to communicate, learnt or intuitive type of quality, which really undercuts the kind of individual knowledge-bearing that is essential to the knowledge argument.
The costs to liberalism of undercutting the idea that individual humans can know things
I now want to turn more generally to the importance to liberalism of the idea that individual humans can know things. Again, I think this can be seen particularly clearly in the central liberal idea of free agency, in that sense of acting of your own accord or your own volition. And on my view, this means something like — this is a kind of Lockean type take — I have the capacity to decide what to do or not to do, and I have the capacity to act on such decisions. Now, these capacities clearly depend on some upstream capacities. In the ordinary sense of these matters, these capacities depend on having mental states — on having thoughts, and so on — on having the capacity for consciousness, on being able to do stuff like think and reason and conclude.
Again, I’m talking here about the very ordinary sense of deciding. I’m not talking about some zombie, thin, sense in which computer scientists and AI people and some philosophers talk about ‘computers deciding’ without meaning this to imply that computers have an internal world or subjective experience. In the ordinary sense, deciding is surely contingent on being able to have beliefs and bear knowledge, at least in the sense of holding beliefs that correspond with some facts about the world.
Think here about the difference between somebody believing in a conspiracy theory, no matter how much evidence is thrown at them to the contrary, and somebody believing something quite ordinary that matches quite a lot of evidence. So, the belief that you were late for work yesterday, or that the glass broke when she dropped it, or that lately fewer people who are coming into my shop are keen to buy strawberry ice cream than they used to be.
It’s this simple sense of being able, at least some of the time, to pick out facts about the world. And, if we take this to be what Hayek is getting at when he refers to knowledge in the knowledge argument, then we don’t need to be able to provide a deeper theory of the relation between belief and knowledge. All we need to know is that without the capacity for that kind of thing, the market would collapse, because knowing things is a large part of what’s going on. The actions of reasoning, concluding, acting, are constitutive of what’s going on, in other words. So, it’s not just that the system reflects the activities of the participants in the market in that its output is prices, it’s that the system is made up of people doing these things.
This combination of capacities is important to liberal arguments beyond economics, of course. For instance, liberals often care about things like respecting rights, holding people to account, and fighting against oppression, and you can’t do any of that without being able to be aware of what’s going on in the world — without coming to some conclusions about stuff, reasoning about these conclusions, and acting on them. And cutting away at this by focusing mostly or solely on tacit knowledge seems very risky for the liberal, particularly within interactions where reasoning and acting on reasons seems essential.
Again, this isn’t to deny that there are good philosophical arguments that we should be skeptical about knowledge — as it happens, I’m pretty much a skeptic about knowledge in the Humean sense — particularly on the grounds that aiming for knowledge, per se, is often too high a bar. Rather, I want to emphasise the tension between the conclusions of those arguments and the standard commitments of liberalism.
The costs to liberalism of promoting the idea that it’s not just individual humans that can know things
Finally, a different risk arises when rather than undercutting the idea that individual humans can know things, instead liberals promote the idea that it’s not just individual humans that can know things. I’m not referring here to beliefs about non-human animals having capacities like knowing and reasoning and acting on their beliefs. Although I do think that the implications of holding those beliefs provokes some pretty serious problems for liberalism — problems that aren’t acknowledged enough. I’m also going to set aside for now the complications for liberalism that arise on the view that non-living things can know stuff — non-living things like abstract systems, or systems made in part of physical things and in part of abstract things. Though again, I do think if you believe that, then you have to reorder some of your liberal thinking.
But even more revolutionary, however, seems to be the idea that groups can have these kinds of capacities. This is because it hits full-on against the individualism that’s at the heart of liberalism. The individual has an undeniably special status within liberalism — whichever kind of liberalism you espouse. There’s a central focus on protecting the individual from the collective, perhaps on seeing individuals as rights-bearers or interest-holders, on individuals as being equally important in themselves. Although again, the wording differs — the commitments differ — between different theories of liberalism.
What’s more, this focus on individualism often implicitly or explicitly tracks the capacities of the individual. So, liberals say things like, “He is capable of making decisions for himself, therefore he should be free to do so,” or, “He chose to do the bad thing to her, so we should hold him to account.” And if groups also have these kinds of capacities that we liberals value and respect in individuals — the capacities to decide and choose, for instance — in this big ordinary sense that’s being referred to in these standard liberal arguments, and again which depend upon things like having thoughts, having beliefs, bearing knowledge — even in just this slightly loose sense in which I’m talking about it, where it’s just something that sort of tracks facts of the world — then how do we reassess our approaches in order to take the interests of the group into account, too? How do we apply our liberal thinking in order to respect and value some non-individuated, non-unified set of things? This seems like a big problem! And I say ‘non-unified’ here because, again, if the group is a ‘single’, then some of the problem goes away.
To that end, I’m going to end by noting two problems that I think are rubbing up against each other here. One is the problem of how the individual humans could come together to collectively know things. This is the kind of ‘how do you make the hive-mind?’ problem. I refer to it as a ‘hive mind’ because ‘group knowing’, in the sense of a group of otherwise individual knowers ‘knowing together’, would surely mean being able to know something because another member of the group knows it. And this implies that you have access to each other’s thoughts, in some sense of sharing thoughts. Maybe even ‘doing thinking’ collaboratively, rather than just like ‘sharing these thoughts’ — actually thinking the things through, coming to something together. But how could this happen? I mean, that’s a great question, but I don’t even need to get on to that here. I can just say things like, what would it mean for our liberal theories of responsibility and respect? How would we cope with it psychologically, as individuated things?
The second different but related problem is the problem of how the ‘non-hive mind group’ could bear knowledge. So, if you want to set aside the hive-mindy stuff, but you still want to say ‘the group bears knowledge’, then if the members of the group cannot in some sense ‘share thoughts’, then what is it that is doing the knowing? What is it that is bearing the knowledge, on the idea that there are things that the group knows that none of its members knows? I mean, it’s either the case, right, that there are things that the group knows that none of its members knows, but they’re kind of sharing them all together, or it’s the case that the group knows the things that none of its members knows, but there is no unified thing to know the stuff. This seems to me very complicated!
These are important questions. I want to work out answers to them. I think my simple answer is just that the group doesn’t know stuff! But I’ll put that to one side because all I want to emphasise for now is that if you buy answers to these problems — such that you have found some clever way to conclude that yes, groups also can bear knowledge in this simple ordinary sense in which individuals can — then you’re going to have to consider whether you should open up any liberal commitments you have to individuals to groups. And you’re going to have to take into account the serious costs of that to the individualism that’s at the heart of liberalism.